What does it mean when AWS says lambda has limit for 1000 concurrent execution?

Question

I am trying to understand the limits of AWS lambda mentioned on the AWS Lambda Limits documentation page.

It says that Lambda has a limit of 1000 concurrent executions .

Let's say we have a Lambda FunctionX , It takes a folderName from an S3 bucket and processes each item in that folder.

Now, I know that Lambda serves each request independently. So if 1000 users try to Invoke FunctionX , with different folder names, FunctionX will serve each request independently and process each folder.

Does this mean if 1001 requests are made then the Lambda FunctionX will not process the additional 1 request?

OR,

By concurrency, does AWS mean, when FunctionX is executing for user1 , it can process 1000 items in that folder concurrently. (Can create 1000 threads to process the items in the folder?)

OR,

I can have N different lambda functions, (F1, F2...., F1000, F1001, FN), handling different functionality. And at a time any 1000 unique Lambda functions can execute concurrently?

Answer 1

That AWS Limits documentation page links to Managing Concurrency - AWS Lambda , which says:

Account Level Concurrent Execution Limit

By default, AWS Lambda limits the total concurrent executions across all functions within a given region to 1000.

So, the answer to your question is that a maximum of 1000 Lambda functions can be executing at any one time in that account .

Further down, it says:

Throttling Behavior

On reaching the concurrency limit associated with a function, any further invocation requests to that function are throttled, ie the invocation doesn't execute your function.

What happens to your extra execution depends on how it was triggered. In some cases, it is rejected. In other cases, it is queued.

For example, if the function is executed synchronously via the Invoke command, it will fail.